Bio-graphics, BioSQL and Rails part 2

In  part 1 of this series we created a rails application and connected it to a BioSQL database. We also overwrote the rails convections to accommodate our legacy schema.

To understand the BioSQL schema, please review the documentation here. A brief overview of is as follows. Every record we enter into our database is a ‘bioentry’ and goes to the bioenty table. A bioentry can be composed of the following entities: the record’s public name, public accession and version, its description and an identifier field.

The actual sequence data is stored in the biosequence table which contains raw sequence information associated with a bioentry, and alphabet information (‘protein’, ‘dna’, ‘rna’). This is because not all records in our database need to be associated with a raw sequence. Additional sequence information is stored in the seqfeature table together with other qualifiers.

The location of each seqfeature (or sub-seqfeature) is defined by a location entity, describing the stop and start coordinates and strand. This information is stored in the location table.

In our rails application we are going to create some models and a few controllers. In RESTful language, we are actually creating resources. In this example we will be very simplistic and just create a biodatabase, taxon, bioentry, biosequence, seqfeature, location resources. We will also create associations between them in their model classes. But before that delete the index.html file from your rails application public folder and add the following line to your configurations/routes.rb file

 map.root :controller => "biosequences"

To quickly create the models, controllers, associated views and a test suite for each of our resources, just run the rails generate scaffold command, passing the name of the model as an argument. For example,

generate scaffold Bioentry

will create a bioentry model, a bioentries_controller, associated views (index,show,edit and new), a migration file, though in our case we do not need it. When you finish scaffolding, the routes.rb file should have the following resources declared.

  map.resources :seqfeatures
  map.resources :locations
  map.resources :bioentries
  map.resources :biosequences
  map.resources :taxon
  map.resources :biodatabases

Let us create some mandatory associations for the models.

Edit the /models/biodatabase.rb file by adding the following

 has_many :bioentries #a biodatabase is associated with many bioentries
 validates_uniqueness_of :name  #The name foe each biodatabase is unique!

Edit the /models/bioentry.rb file by adding the following

    belongs_to :biodatabase
    belongs_to :taxon
    has_one :biosequence

Edit the /models/taxon.rb and add

   has_one :bioentry

Edit the /models/biosequence.rb file by adding:

  set_primary_key :bioentry_id #biosequence uses bioentry_id as a primary key!
  belongs_to :bioentry

edit the /models/location.rb file by adding:

 belongs_to :seqfeature

Edit the /models/seqfeature.rb file by adding:

  belongs_to :bioentry
  has_many :locations

Note that most likely you will be adding huge files to the database. BioSQL comes with a set of  perl scripts to enable you do that. Until bioruby 1.3 is released you will have to use the perl scripts to add huge datasets. All the documentation to do that is available from the BioSQL website. I used a perl script load_ncbi_taxonomy.pl to load taxon data to my database. This script comes with the BioSQL. (It did not seem to work on my system, I will sort that later)

To make this post shorter and get to the meat of it, i will assume that you have some existing data in your biosql database. If not, create some dummy data to populate, the biodatabase, bioentry,biosequence, seqfeature and location tables. In Part 3, I will show you how to create the necessary views to populate the database. After all biologists don’t want to interact with raw SQL queries and sometimes have no idea of running scripts, however they are very web savy!

Edit the /biosequences/show.html.erb to look as follows:

<h2><%= @biosequence.bioentry.name%>(<%= @biosequence.alphabet %>)</h2>
<p>Sequence</p>
<%= @biosequence.seq %><br/>


<%= link_to 'Edit', edit_biosequence_path(@biosequence) %> 

Now navigate to http://localhost:3000/biosequences/1

and then navigate to http://locahost:3000/biosequences/1.xml The XML version of your sequence is also available!

Lets add some ability to render graphics for the sequences.

Add the following lines at the top of the biosequence.rb model file

 require 'stringio'
 require 'base64' 

In the biosequence.rb model class, create a new method called draw_graphic.

def self.draw_graphic(value)
      #get the name and length of the main feature to be drawn
     main_feature = Bioentry.find(value)
     len = main_feature.biosequence.length.to_i
     name = main_feature.name

    #create a Biographics panel and add a track
      @my_panel = Bio::Graphics::Panel.new(len,:width=> 900)
      @track = @my_panel.add_track("#{name}",:glyph=>'directed_generic')

     #specify the range for the main feature
     main_feature_range = "1..#{len}"
      @track.add_feature(Bio::Feature.new("#{name}",main_feature_range), :label=>" ")

    #write the output to memory
        output = StringIO.new
        @my_panel.draw(output)
        return output.string
  end

This method will be called by an action method in biosequence_controller.rb file.

  def to_image
    begin
      image = Biosequence.draw_graphic(Biosequence.find(params[:id]))
      send_data(image, :filename => "graphic.svg", :disposition => "inline")
    rescue  ActiveRecord::RecordNotFound
      add_error("Error:Attempt to call image without specifying a biosequence  ID")
      redirect_to :action=>'index'
    end
  end

We add a rescue block to capture record not found errors. In RESTful applications a controller is limited to seven actions. So we need to add a collection to our biosequence resource in routes.rb. This is how we do it.

  map.resources :biosequences,:collection=>{:to_image=>:get}

Now we need to modify our /biosequences/show.html.erb file, to enable rendering of the graphic. For that we will create a helper method so that our show.html.erb view is ‘clean’. In helpers/biosequences_helper.rb file, add the following code

  def render_image(feature_obj)
     image_tag(url_for({:action=>'to_image',:id=>feature_obj}))
  end

And in the /views/biosequences/show.html.erb file add the following line of code

<%= render_image(@biosequence) %><br/>

Now assuming  that you have a biosql database with valid data, navigate to

http://localhost:3000/biosequences/show/1

screenshort

screenshort

The above is a screen shot from my example application while I was writing this tutorial.

The source code for this example  application is available from github

For a full review of the methods available for biographics please check the project’s git repository and the rdoc.

Bio-graphics, BioSQL and Rails part 1

In these series I will show you how to quickly add graphics support to a bioinformatics database rails application. We are going to use the biographics library by Jan Aerts, the BioSQL database schema, and rails 2.2.2 (also works with 2.3.2)  In this simple example we want to represent a sequence as a graphic, such that we can view it in a web browser more or less the way Gbrowse works. Each main feature has different subfeatures at different locations along it.

——————————————————————- main feature

——- ——– ———— ——– ——    subfeatures

We need to have the following installed, rails 2.1.1, bio 2.1, biographics 1.4 all available as gems and a database based on BioSQL schema.

We need to download the BioSQL schema located here. The latest version as of this writing is BioSQL v1.0 (code-named Tokyo) release, v1.0.1. Create a database called biosql_development. I am on Ubuntu Linux with Mysql 5.0.

george:>mysql -u george -p
:enter password
Welcome to the MySQL monitor.  Commands end with ; or \g.
Your MySQL connection id is 27
Server version: 5.0.51a-3ubuntu5.4 (Ubuntu)

Type 'help;' or '\h' for help. Type '\c' to clear the buffer.

mysql> create database biosql_development;
Query OK, 1 row affected (0.02 sec)

mysql> 

I have created a database called biosql_development. Why am i not using migrations? The reason is that BioSQL has some agreed standards on table names and schema convection which are not compatible with rails database creation and table naming conventions. However Rails allows us to override these default convections, when working with legacy databases, as will be our case.

After creating the database, load the BioSQL schema to the empty database. First we need to tell mysql which database to use.

mysql> use biosql_development;

then load the schema

mysql> source /home/george/Desktop/downloadsfolder/biosql-1.0.1/sql/biosqldb-mysql.sql;
Query OK, 0 rows affected, 1 warning (0.48 sec)

Query OK, 0 rows affected (0.15 sec)
Records: 0  Duplicates: 0  Warnings: 0

Query OK, 0 rows affected, 1 warning (0.01 sec)

Query OK, 0 rows affected (0.03 sec)
Records: 0  Duplicates: 0  Warnings: 0

Query OK, 0 rows affected, 1 warning (0.01 sec)

 ........ trucated
mysql>

Now we need to create a rails application and connect to this database.

I use the Netbeans IDE development environment for creating ruby and rails applications. Go ahead and create a rails application and specify to use mysql as the database adapter.

To connect to our legacy database, we need to override some convections. First disable table plurulization, and tell rails that the table primary name is named as tablename_id as opposed to just the id column expected by rails. To do that

Create a new file in your application configurations/initializers directory called override_rails.rb (you can call it whatever).

 class ActiveRecord::Base
  self.pluralize_table_names = false

  self.primary_key_prefix_type = :table_name_with_underscore
 end

The two lines above tells ActiveRecord not to expect the table names to be plural and that the primary key for each table is named as tablename_id format.

Also create another one called external_libraries.rb in the initializers directory, as you can tell this is where I want to put my require statements for loading external libraries.

require 'rubygems'

#load the bioinformatics library
require 'bio'

#load the biographics library
require 'bio-graphics'

#load the sql views extension library
gem 'rails_sql_views'
require 'rails_sql_views'

This file loads our gems. The rails_sql_views gem allows us to create views and access them by creating models corresponding to the views.

At this point if you run rake db:schema:dump, we will have a rails based BioSQL schema and which we can conveniently use to create a BioSQL database on any Relational database that rails supports and this includes Microsoft SQL server, DB2, Oracle, SQLlite and a host of others. All that would be required is to change the database.yml file to suit the adapter of choice and then execute rake db:schema:load to load the BioSQL schema.

Please note that if your are using rails 2.2.2,  you may want to comment the lines

unless Kernel.respond_to?(:gem)
  Kernel.send :alias_method, :gem, :require_gem
end

in rails_sql_views(0.6.1), otherwise running db:schema:dump will cause rake to abort.

In the next part I will describe how to create the necessary resources for our RESTful(Representational State Transfer) bioinformatics web application and rendering of the graphics.


Approximate string matching metrics with amatch

Most often in sequence analysis we want to compare how  similar two sequences are. How can we quantify similarity by using a metric? That was my question yesterday and I went hunting for a ruby implementation for such metrics. Luckily I got a library called amatch which is an approximate string matching extension for ruby! amatch implements the following metrics:

Hamming distance, Levenshtein edit distance,longest subsequence common to two strings,longest substring common to two strings,sellers distance and pair distance which is based on the number of adjacent character pairs, that are contained in two  strings.

Hamming distance

This is the number of characters that are different between two strings. This is not recommended for the majority of string based information retrieval. Very similar strings can sometimes be given high hamming distances.

Leveshtein edit distance

Is defined as the minimal costs involved in transforming one string into another by using  deletion, insertion and substitution of a character to one of the strings. The algorithm can associate a cost for performing each of the operations and for this metric it is usually 1.

Longest common substring

This is define as the contiguous chain of characters that exists in both strings. The longer the substring the better the match between the two strings. The problem with this approach is that if a difference was introduced in the middle of one string, the distance will be longer that if the same difference was introduced at the beginning of one of the strings.

Longest common Subsequence

The longer the common sub sequence is, the more similar the two strings will be. In this case a sub sequence does not have to be contiguous.

Look at the documentation for more explanations of the metrics and algorithms.

To use the library you need to first install the gem. I installed it on my Linux box running Ubuntu and ruby 1.8.6.

sudo gem install amatch

Then in script,

require 'rubygems'

require 'amatch'
include Amatch
require  'bio'
#with bioruby it would be easy to compare two sequence entries  for example
seq_obj1 = Bio::Sequence.auto("actagatatttgat")
seq_obj2 = Bio::Sequence.auto("gccagatagttaat")

#calculate the hamming distance
 m = Hamming.new(seq_obj1.to_seq)
 m.match(seq_obj2.to_seq)
#=> 

#calculate pair-distances between the two sequences
pair_distance_obj = PairDistance.new(seq_obj1.seq)
pair_distance_obj.match(seq_obj2.seq)
 #=>
# note that you can just substitute the strings directly to the metric object creation method
without creating the sequence objects!

Note that amatch  failed to install on windows XP with the following error

Building native extensions.  This could take a while…
ERROR:  Error installing amatch:
ERROR: Failed to build gem native extension.

C:/ruby-1.8.6/ruby/bin/ruby.exe extconf.rb install amatch
creating Makefile

nmake
‘nmake’ is not recognized as an internal or external command,
operable program or batch file.

Although i have nmake installed on my windows machine. I will look at that later.

Happy string matching!


What’s new in Bioruby edge

Changes to Bio::Blast

Naohisa Goto has announced changes to the Bio::Blast.reports to support default -m 0 and tabular -m 8
formats in addition to XML (-m 7) form. I think this is really nice and convenient!

Previously it meant that for bioruby to parse a Blast file, you had to have your blast results in XML output which Bio::Blast::Reports would understand. However, by default Blast gives an  -m0 output and without that prior knowledge you may spend hours wondering what is wrong when parsing default blast output files. With Bioruby 1.2.1 this will not work

require 'rubygems'
require 'bio'

report_file = "/home/george/esther_blast_files/blast_output2.txt"
Bio::Blast.reports(report_file) do |report|
  puts report.class
end

Unless blast_output2.txt is in XML format

In the upcoming Bioruby 1.3 the default Blast file can now be parsed, for example,

require 'rubygems'
require 'bio'
report_file = "/home/george/esther_blast_files/blast_output2.txt"
Bio::FlatFile.open(Bio::Blast::Default::Report,report_file) do |ff|
ff.each do |rep|
   puts rep.statistics
   rep.iterations.each do |itr|
      puts itr.hits.size

    itr.hits.each_with_index do |hit,i|
      puts hit.hit_id
      puts hit.len
     end
   end
 end
end

Bio::Blast.remote now supports DDBJ in addition to Genomenet. It would be a nice idea to support NCBI as well.

Changes to Bio:sequence

It is possible to create  sequence objects from Bio::GenBank, Bio::EMBL, and Bio::FastaFormat by using the to_biosequence method

gb = Bio::GenBank.new(genbank_file.gb)
gb.to_biosequence

Bio::SQL Support

Thanks to Raoul and Naohisa, support for BioSQL has been rewritten by using  ActiveRecord.

#Make a connection

connection = Bio::SQL.establish_connection(path_to_database.yaml,'development')
#list databases

databases =Bio::SQL.list_databases

#retrieve a sequence
sample_seq = Bio::SQL.fetch_accession('some_accession_number')

#get number of seqeunces in the database
puts Bio::SQL.list_entries

#get references associated with an entry
puts sample_seq.references

#create an embl format
puts sample_seq.to_biosequence.output(:embl)

Changes to Bio::GFF2 and Bio::GFF3

GFF2/GFF3 formatted texts are now supported but there will be backward portability issues with bio 1.2.1 since some incompatible changes have been incorporated.  Bio::GFF::Record.comments has been renamed to comment and comments= is now comment=

Both Bio::GFF::GFF2::Record.new and Bio::GFF::GFF3::Records.new, can now take 9 arguments that correspond to GFF columns making it easy to create a Record object directly without need for  formatted text.

Both Bio::GFF::GFF2::Record#attributes and Bio::GFF::GFF3::Record#attributes have been changed to return a nested array containing tag, value pairs, to obtain a  hash, use the to_hash method

To support data output for GFF2/GFF3, new methods have been added:  Bio::GFF::GFF2#to_s, Bio::GFF::GFF3#to_s, Bio::GFF::GFF2::Record#to_s,and Bio::GFF::GFF3::Record#to_s

Lots of other changes have been incorporated for the GFF classes and you can view the change log at github

CodeML parser

A wrapper for PAML codeml program that is used for estimating evolutinary rate has been added.  The class provides methods  for generating the necessary configuration file. The new Bio::PAML::Codeml::Report and PAML::Codeml::Rates  provides simpel classes  for accessing the codeml report and rates file.  This example is from the example given in the source code

require 'bio'
# Reads multi-fasta formatted file and gets a Bio::Alignment object.
     alignment = Bio::FlatFile.open(Bio::Alignment::MultiFastaFormat, 'example.fst').alignment
     # Reads newick tree from a file
     tree = Bio::FlatFile.open(Bio::Newick, 'example.tree').tree
  # Creates a Codeml object
  codeml = Bio::PAML::Codeml.new
     # Sets parameters
     codeml.parameters[:runmode] = 0
     codeml.parameters[:RateAncestor] = 1
     # You can also set many parameters at a time.
   codeml.parameters.update({ :alpha => 0.5, :fix_alpha => 0 })
     # Executes codeml with the alignment and the tree
     report = codeml.query(alignment, tree)

Lots of Bugs  have been fixed and also support for Ruby 1.9 has been added. Its great thanks to the bioruby developers for their time and the excellent new changes!



Bioruby mini-series: The Bio::Sequence::Common class

Sequence Transformation

Lets have a look at the Bio::Sequence::Common class module which provides us with most of the sequence transformation methods for biological sequences.

Bio::Sequence::Common

Implements methods which are common to both Bio::Sequence::AA and Bio::Sequence::NA, for example

A Bio::Sequence object is easily created like this;

require ‘bio’

my_dna = Bio::Sequence.auto("actagatatttgat") #=> actagatatttgat

my_dna is now a Bio::sequence object and you can use the various methods available for this class, which we are going to explore shortly.

Bio::Sequence::Common Non Modifying methods

  • to_s

    This method returns a sequence as a string. It does not modify the original sequence.

    puts my_dna.to_s #=> actagatatttgat

    puts my_dna.to_s.class #=> String

    An alias for this method is the to_str method.

    my_dna.to_str
    #=> actagatatttgat

  • seq

    This method will return a new Bio::Sequence::NA or Bio::Sequence::AA object. The original sequence remains unchanged. For example if you wished to assign a new instance of my_dna object that we created above ,such that you have a my_dna2 object, you would create that as follows,

    my_dna2 = my_dna.seq

    puts my_dna2 #=> actagatatttgat

    puts my_dna2.class #=>
    Bio::Sequence::NA

Bio::Sequence::Common modifying methods

  • Normalize!

    This method removes all the white space and transforms all positions to uppercase if the sequence is an amino acid (AA) or transforms all positions to lowercase if the sequence is a nucleic acid (NA) sequence, leaving the original sequence modified

    For example

    test_seq = Bio::Sequence::NA.new(“ACTG”)

    puts test_seq.normalize! #=>
    actg

  • Concatenating

    Many times we want to append a new sequence or a set of bases/residues eg a poly A sequence to the end of a new sequence and modify the original sequence. This is achieved by the concat method.
    It is also referred to as << method.

    test_seq = Bio::Sequence::NA.new(“actg”)

    test_seq << “acagat”

    test_seq concat “acagat”

    puts test_seq #=>
    actgacagat

Note that to create a new sequence that adds to an existing sequence without altering the original sequence you would use the + operator. It accepts a variable number of arguments. For example

test_seq = Bio::Sequence::NA.new(“actg”)

test_seq2 = test_seq + (“cttcccttttt” “tatatata”)

puts test_seq2 #=>
actgcttcccttttttatatata

puts test_seq #=>actg

Working with subsequences

Please note that biological sequence numbering convections are one based as opposed to ruby’s zero based. Biological coordinate’s convection for BioSQL and Chado is zero based.

  • Subseq

    This method returns a new sequence containing the subsequence identified by the start and end values given as parameters. This method works in a similar way to the slice string method. For example

    my_seq = Bio::Sequence::NA.new(“agggatttc”)

    puts my_seq.subseq(2,5) #=>
    ggga

    The first argument denotes the start and the second argument denotes the end of the subsequence. Both arguments must be positive integers

    When this method is used without arguments, the start defaults to 1 and the end defaults to the last element of the string. Therefore when subseq is called without any arguments, it returns a new sequence similar to the original sequence.

    puts my_seq.subseq #=> agggatttc

  • window_search

    This method is typically used with a block. The method is called if you wanted to step through a sequence given a length of a subsequence. Therefore the method accepts two arguments. Step_size which defines the size of your ’steps’ and the window_size which defines the length of the stepping subsequence. Any remaining sequence at the terminal end will be returned. The default step size is one since its an optional argument.

    For example

    To print the average GC% on each 100bp you can write,

    s.window_search(100) do |subseq|

    puts subseq.gc

    end

Bioruby mini-series: The Sequence class

Bioruby is a bioinformatics ruby package for analysis of biological sequences. In my quest to become a bioruby guru i have decided to poke the bioruby API and all available tutorials to better understand this fantastic library written by the bioruby team of developers. My journey will be logged here as the bioruby mini series. We start with an introductory overview of the sequence class.

To use the library you need to have a ruby interpreter installed , preferably ruby 1.8.5 and above . To install bioruby as a gem, do:
sudo gem install bio

This will install Bioruby version 1.1.0 and it comes with its own shell as well.

Type bioruby on the command prompt and you will see this:


Loading config (/.bioruby/shell/session/config) … done
Loading object (/.bioruby/shell/session/object) … done
Loading history (/.bioruby/shell/session/history) … done

. . . B i o R u b y i n t h e s h e l l . . .

Version : BioRuby 1.1.0 / Ruby 1.8.6

bioruby>

Now we ready to rock and roll! I dug in to the API and extracted some useful information for us.

The Bio::Sequence class

This is the primary sequence class and deals with sequence translation and transformations. It inherits from ruby’s string class which means that you can use ruby’s string methods with the Bio::Sequence class just like you would with a string.

The Bio::Sequence class object is a wrapper around the actual sequence and it is represented as either a Bio::Sequence::NA or a Bio::Sequence::AA. and responds to all the methods that are defined for both NA and AA classes. This class has the following methods:

  • auto – This will guess the type of sequence provided and return the appropriate Bio:Sequence class for the given string, either a Bio::Sequence::AA or a Bio::Sequence::NA
  • new – Creates a new Bio::Sequence object. It does not initialize the object in to any of the bioruby objects. It returns a string.
  • aa – Will transform your current Bio::Sequence object to a Bio::Sequence::AA object. It will change your current object i.e it will transform a Bio::Sequence::NA to a Bio::Sequence::AA which is undesirable. So it needs to be used only when you are sure of the type of sequence you are working with.
  • na – works the same as the aa method above but the returned object is a Bio::Sequence::NA
  • output – It returns a string with the current Bio::Sequence object formatted with the given style. The supported styles are fasta, genbank and embl. The style argument is passed as a ruby symbol eg :fasta
  • to_s – it returns the sequence as a string leaving the original sequence unaltered. The to_str is an alias for this method

Bio::Sequence::NA class

This class wraps a nucleic acid sequence. It provides a number of methods to work with a DNA sequence as demonstrated in the example below.

Dr Optimist has finally finished his long awaited sequencing project code named Sikwensi. The nucleic acid sequence for a chromosome for which he won’t reveal any further details is shown below.
“gacagatggacatggactagagctgct”

He calls his trusted ruby programmer to help analyze the sequence and tear it base by base. The guy gets to work.

require ‘bio’

bio_seq = Bio::Sequence.auto( ‘gacagatggacatggactagagctgct’) #=> bio_seq is now a Bio::Sequence::NA object

#get the number of codons in the sequence

bio_seq.window_search(3,3) {|codon| puts codon}

# complemental sequence

bio_seq.complement (Bio::Sequence::NA object)

# gets subsequence of positions 4 to 14
bio_seq.subseq(4,14) # he thinks the subsequence is interesting and worth
extracting!

bio_seq.gc_percent #what is the gc content?

bio_seq.composition # nucleic acid compositions (returns a Hash)

bio_seq.translate # translation ( returns a Bio::Sequence::AA object)
bio_seq.translate(2) # translation from frame 2 (The default is frame 1)
bio_seq.translate(1,11) # using codon table No.11 (bacteria)
bio_seq.translate.codes # shows three-letter codes ( returns an Array)
bio_seq.translate.names # shows amino acid names (returns an Array)
bio_seq.translate.composition # amino acid compositions (returns a Hash)
bio_seq.translate.molecular_weight # calculating molecular weight (returns Float)

bio_seq.complement.translate # translation of complemental strand

A tutorial written by Katayama Toshiaki can be found here and translated to English by Naohisa Goto. (Thank you guys!)